Composite selection signals: application in purebred dogs

Abstract

Composite selection signals (CSS) have demonstrated that the locus or interesting regions can be localised in multi-breed populations. Here, we investigated the application of CSS to canine SNP data to assess previously known signatures or identify novel regions from various purebred dogs breed comparisons. Chapter 2 was to apply CSS to a publicly available dataset to construct a high-resolution positive selection map for three complex traits in Asian dogs. Estimates of CSS provide clear evidence that any regions of significant SNPs captured is highly likely to contain a strong candidate genes or variants of positive selection via the breed grouping strategy and methodology. Next, Chapter 3 was to identify genomic region under positive selection in Bullmastiffs. Most notable were genomic regions, candidate genes and related pathways associated with the distinct morphology of the breed. Chapter 4 was to identify selection signatures in the Mastiff-like breeds that correspond to lymphoma susceptibility. The combination of phenotypes based on breed-specific disease risk profiles with CSS has been successful in defining genomic regions that may contribute to lymphoma risk in Mastiff-like dog breeds. Finally, Chapter 5 presents of analyses CNV and selection signatures in a population of Australian Border Collies, including a sub-group of lymphoma cases. The use of the CSS approach for the analysis of lymphoma susceptibility based on positively selected regions in Border Collies produced a range of novel gene candidates that provide a basis for investigation of the genomic landscape of canine lymphoma in future studies. Overall, the methodology used in this study provides a detailed investigation about the core traits influenced by the strong selection events in canine breeds and presents novel insights about the hotspots of positive selection in the canine genome.